Mathematical Reasoning on IneqMath 1252/100 (dev)
14Bound AccuracyOff-the-Shelf
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| Off-the-ShelfModel=Llama3.2-1B-Instruct, Protocol=Zero-shot, K=02025.10 | 14 | 12 | 13 | |
| MHA-ESPModel=Llama3.2-3B-Instruct, Protocol=MHA-ESP Retrieval, K=22025.10 | 14 | 22 | 18 | |
| LoRAModel=Llama3.2-1B-Instruct, Protocol=Fine-tuned, K=22025.10 | 12 | 26 | 19 | |
| MHA-ESPModel=Llama3.2-1B-Instruct, Protocol=MHA-ESP Retrieval, K=22025.10 | 12 | 20 | 16 | |
| LoRAModel=Llama3.2-3B-Instruct, Protocol=Fine-tuned, K=22025.10 | 12 | 26 | 19 | |
| Off-the-ShelfModel=Llama3.2-3B-Instruct, Protocol=Zero-shot, K=02025.10 | 10 | 22 | 16 | |
| ICLModel=Llama3.2-3B-Instruct, Protocol=In-Context Learning, K=22025.10 | 10 | 24 | 17 | |
| ICLModel=Llama3.2-1B-Instruct, Protocol=In-Context Learning, K=22025.10 | 4 | 16 | 10 |